Abstract No. 619 Percutaneous transhepatic obliteration of bleeding rectal varices utilizing sodium tetradecyl sulfate foam and interlocking detachable coils: a single-center experience

Abstract

To assess technical predictors of Transjugular Intrahepatic Portosystemic Shunt (TIPS) dysfunction. Materials: A retrospective cohort analysis of 235 consecutive patients that underwent TIPS placement from Jan 2014-Feb 2017 was analyzed. Technical factors assessed included TIPS location (proximal and distal landing zone), operator experience, portal vein thrombosis, TIPS angle acuity, and stent length. TIPS angle was defined as the angle of stent entry into the portal vein and was measured by two independent radiologists blinded from the analysis. Proximal landing zone was defined by shunt origin in the inferior vena cava (IVC), right hepatic vein (RHV), middle hepatic vein (MHV), or left hepatic vein (LHV). Distal landing zone was categorized as right, left, or main portal vein (RPV, LPV, MPV). A multiple logistic regression model was used for analysis. Results: Of 235 patients who had initial TIPS placement, 17.4% (41/235) underwent subsequent revision for TIPS dysfunction. TIPS dysfunction was due to stenosis at the hepatic venous end (24/41), mid-portion of the stent (2/41), portal venous end (6/41), complete thrombosis (6/41), and multifactorial in (3/41). Median follow-up was 27.1 mo. Average time after TIPS placement to TIPS dysfunction was 5.7 ± 5.0 mo. TIPS was most commonly placed in RHV extending to RPV (68.5%, 161/235), followed by MHV to RPV (11.9%, 28/235), with the rest originating in MHV or IVC and ending in MPV or LPV (19.5%, 46/235). Decreasing angle of entry into the portal vein was associated with greater likelihood of TIPS dysfunction (OR 0.926, 95%CI .904-.949, p<.001). Average TIPS angle for patients with TIPS dysfunction was 111.8 ± 19.1 compared to 136.5 ± 15.6 in those without dysfunction (p<.001). Stent length (p ¼ .304), operator experience (p ¼ .344), portal vein thrombosis (p ¼ .388), proximal TIPS landing zone (p ¼ .324), and distal TIPS landing zone (p ¼ .167) were not predictors of TIPS dysfunction. Conclusions: The angle of TIPS stent entry into the portal vein is a predictor of future TIPS dysfunction. Stent length, shunt location, operator experience, and portal vein thrombosis were not associated with TIPS dysfunction.